1
2
3
4
5
Index(['uid', 'oil_actv_dt', 'create_dt', 'total_oil_cnt', 'pay_amount_total',
       'class_new', 'bad_ind', 'oil_amount', 'discount_amount', 'sale_amount',
       'amount', 'pay_amount', 'coupon_amount', 'payment_coupon_amount',
       'channel_code', 'oil_code', 'scene', 'source_app', 'call_source'],
      dtype='object')

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
def time_isna(x,y):
    if str(x) == 'NaT':
        x = y
    else:
        x = x
    return x
df2 = df.sort_values(['uid','create_dt'],ascending = False) 
# 按照uid和create_dt降序排列->感觉这一步没什么意义？
df2['create_dt'] = df2.apply(lambda x: time_isna(x.create_dt,x.oil_actv_dt),axis = 1)
# axis = 1就是对行进行循环，对每一行使用apply里面的函数

# 做补全的时候，也可以考虑下面的方法：
# df2_test = df.sort_values(['uid','create_dt'],ascending = False) 
# df2_test.loc[df2_test.create_dt.isna(), 'create_dt'] = df2_test.loc[df2_test.create_dt.isna(), 'oil_actv_dt']

df2['dtn'] = (df2.oil_actv_dt - df2.create_dt).apply(lambda x :x.days)
df = df2[df2['dtn']<180]
df.head()
# 保留申请贷款前180天内的数据

base = df[org_lst]
base['dtn'] = df['dtn']
base = base.sort_values(['uid','create_dt'],ascending = False)
# 按create_dt进行降序排列
# base = base.drop_duplicates(['uid'],keep = 'first')
base = base.drop_duplicates(['uid'],keep = 'last')
# 私以为，求最大间隔天数的话，应该keep last（因为sort_values用的依据的使create_dt）
base.shape

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
gn = pd.DataFrame()
for i in agg_lst:
#     计数
    tp = pd.DataFrame(df.groupby('uid').apply(lambda df:len(df[i])).reset_index())
    tp.columns = ['uid',i + '_cnt']
    if gn.empty == True:
        gn = tp
    else:
        gn = pd.merge(gn,tp,on = 'uid',how = 'left')
    
#     大于零的个数
    tp = pd.DataFrame(df.groupby('uid').apply(lambda df:np.where(df[i]>0,1,0).sum()).reset_index())
    tp.columns = ['uid',i + '_num']
    if gn.empty == True:
        gn = tp
    else:
        gn = pd.merge(gn,tp,on = 'uid',how = 'left')
        
#     非空元素的和
    tp = pd.DataFrame(df.groupby('uid').apply(lambda df:np.nansum(df[i])).reset_index())
    tp.columns = ['uid',i + '_tot']
    if gn.empty == True:
        gn = tp
    else:
        gn = pd.merge(gn,tp,on = 'uid',how = 'left')
        
#     非空元素的平均值
    tp = pd.DataFrame(df.groupby('uid').apply(lambda df:np.nanmean(df[i])).reset_index())
    tp.columns = ['uid',i + '_avg']
    if gn.empty == True:
        gn = tp
    else:
        gn = pd.merge(gn,tp,on = 'uid',how = 'left')
        
#     非空元素的最大值
    tp = pd.DataFrame(df.groupby('uid').apply(lambda df:np.nanmax(df[i])).reset_index())
    tp.columns = ['uid',i + '_max']
    if gn.empty == True:
        gn = tp
    else:
        gn = pd.merge(gn,tp,on = 'uid',how = 'left')
        
#     最小值
    tp = pd.DataFrame(df.groupby('uid').apply(lambda df:np.nanmin(df[i])).reset_index())
    tp.columns = ['uid',i + '_min']
    if gn.empty == True:
        gn = tp
    else:
        gn = pd.merge(gn,tp,on = 'uid',how = 'left')
        
#     方差
    tp = pd.DataFrame(df.groupby('uid').apply(lambda df:np.nanvar(df[i])).reset_index())
    tp.columns = ['uid',i + '_var']
    if gn.empty == True:
        gn = tp
    else:
        gn = pd.merge(gn,tp,on = 'uid',how = 'left')
    
#     最大值-最小值
    tp = pd.DataFrame(df.groupby('uid').apply(lambda df:np.nanmax(df[i]) -np.nanmin(df[i]) ).reset_index())
    tp.columns = ['uid',i + '_var']
    if gn.empty == True:
        gn = tp
    else:
        gn = pd.merge(gn,tp,on = 'uid',how = 'left')
    
#     变异系数
    tp = pd.DataFrame(df.groupby('uid').apply(lambda df:np.nanvar(df[i])/max(np.nanmean(df[i]),1)).reset_index())
    tp.columns = ['uid',i + '_var']
    if gn.empty == True:
        gn = tp
    else:
        gn = pd.merge(gn,tp,on = 'uid',how = 'left')

1
2
3
4
5
6
7
8
gc = pd.DataFrame()
for i in dstc_lst:
    tp = pd.DataFrame(df.groupby('uid').apply(lambda df: len(set(df[i]))).reset_index())
    tp.columns = ['uid',i + '_dstc']
    if gc.empty == True:
        gc = tp
    else:
        gc = pd.merge(gc,tp,on = 'uid',how = 'left')

1
2
3
4
5
fn = pd.merge(base,gn,on= 'uid')
fn = pd.merge(fn,gc,on= 'uid') 
fn.shape

fn = fn.fillna(0)

1
2
3
4
5
6
7
x = fn.drop(['uid','oil_actv_dt','create_dt','bad_ind','class_new'],axis = 1)
# 删去一些列
y = fn.bad_ind.copy()
from sklearn import tree

dtree = tree.DecisionTreeRegressor(max_depth = 2,min_samples_leaf = 500,min_samples_split = 5000)
dtree = dtree.fit(x,y)

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
import pydotplus 
from IPython.display import Image
from sklearn.externals.six import StringIO
import os
os.environ["PATH"] += os.pathsep + 'D:setupGraphvizbin' # 为了找到graphviz
with open( "dt.dot", "w") as f:
    tree.export_graphviz(dtree, out_file=f) # 可视化函数
dot_data = StringIO() # 在内存中读写str
tree.export_graphviz(dtree, out_file=dot_data,  
                         feature_names=x.columns,
                         class_names=['bad_ind'],
                         filled=True, rounded=True,
                         special_characters=True)
graph = pydotplus.graph_from_dot_data(dot_data.getvalue())  # getvalue()方法用于获得写入的str
Image(graph.create_png())